Bridge Model Mapping for Fast Simulator Parameter Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The process of acquiring appropriate parameter values for simulators is lengthy due to the complexity of simulating analysis targets, making it time-consuming for accurate simulation and analysis.
Innovation Solution
A model generation device and method that generates a bridge model to relate parameters between a simulator model and a machine learning model, allowing for the calculation of simulator model parameters using machine learning predictions, thereby reducing the time required for parameter acquisition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional simulation methods are used to acquire parameter values for analyzing analysis targets, then accurate simulation results can be obtained, but the processing time becomes excessively long
Solution Approach 1:
The patent introduces a bridge model as an intermediary between the simulator model and the machine learning model. This bridge model learns the mapping relationship between the two models, enabling parameter values to be transferred from the machine learning model to the simulator model without requiring lengthy traditional simulation processes. The bridge model acts as a mediator that connects the fast but less interpretable machine learning model with the accurate but time-consuming simulator model.
Solution Approach 2:
The bridge model is trained in advance using paired data from both the simulator model and the machine learning model. This preliminary training establishes the mapping relationship before actual parameter acquisition is needed. When parameter values are required, the pre-trained bridge model can quickly transform machine learning predictions into simulator-compatible parameters without requiring the full simulation process to run each time.
2Measurement precision
If multiple complex processes are used to acquire appropriate parameter values, then accurate simulation can be achieved, but the complexity of the process increases
Solution Approach 1:
The patent merges the functionality of multiple separate processes (simulator execution, machine learning prediction, and parameter mapping) into a unified framework. The bridge model consolidates the parameter transformation logic that would otherwise require separate processing steps, combining the strengths of both simulator-based accuracy and machine learning efficiency into a single integrated system.
Solution Approach 2:
The bridge model creates a simplified copy or representation of the relationship between the simulator model and machine learning model. Instead of requiring the full complexity of the simulator to run for each parameter acquisition, the bridge model captures the essential mapping relationship in a lighter-weight structure that can be applied repeatedly without the full simulation overhead.
Data Source
AI summary
A model generation device includes a model generation unit that generates a third model indicating a relationship between a first model and a parameter of a second model, the first model indicating a relationship between a sample and a label of the sample, the second model indicating the relationship and being different from the first model.